An important application scenario of mobile opportunistic networks is social contact networks, in which nodes are composed of mobile devices carried by human beings, and social relations among nodes should be considered when we design routing algorithm for mobile opportunistic networks. In this paper, we focus on social relations among nodes and node movement prediction, and propose a novel social communitybased routing algorithm. We introduce the social relations as the key metrics for

more »

... y dividing, and semi-Markov process is used to model movement of nodes based on the community structure. Based on the result of prediction, the next-hop nodes can be appropriately chose for message forwarding. Simulation is done, and the result shows that the proposed algorithm outperforms the other three classic routing algorithms.